A series of population studies and animal experiments suggests that salt (sodium) and nitrates in food may increase human risk for cancer of the stomach and other digestive tract sites. However, few long-term epidemiological studies have been conducted that measure dietary intake of these substances in relation to incidence of the disease. The self-administered food frequency questionnaire (FFQ) represents an effective and inexpensive approach for collecting dietary information on large cohorts over time. An evaluation of the validity and reproducibility of a 116-item FFQ for measuring sodium, potassium and nitrate is proposed. The main hypothesis to be tested is that the FFQ is an accurate and reliable instrument for measuring sodium, potassium and and nitrate in epidemiologic research. In addition, it is hypothesized that: A. Overnight excretion rates of these ions may significantly underestimate intake; B. Dietary fiber may affect the rate and level of urinary sodium excretion, and influence the association with intake; To test these hypotheses, 350 adult males and females residing or working in Greater Boston will be enrolled into the study. Each subject will provide two 24-hour urine samples, divided into 12-hour segments, and spread six months apart. A FFQ will be completed with both collections, to assess dietary intake of these food constituents over the previous year. The correlations between intake measured by the FFQ and urinary excretion levels will be analyzed using crude and stratified analyses. Based on preliminary data, the design of the study predicts that an observed correlation for sodium and nitrate of .20 and a "true" coefficient of .40 will be detected.